# mmFlux: Crowd Flow Analytics with Commodity mmWave MIMO Radar

arXiv: https://arxiv.org/abs/2507.07331 · HTML: https://mlsanity.com/paper/2507.07331/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 116 references that are in the map.
- First posted on arXiv 9 Jul 2025.

## Builds on (most cited first, 13 of 13)

- [FlowFormer++: Masked Cost Volume Autoencoding for Pretraining Optical Flow Estimation](https://mlsanity.com/paper/2303.01237/) (2023)
- [Context-Aware Trajectory Prediction](https://mlsanity.com/paper/1705.02503/) (2017)
- [Self-Supervised Scene Flow Estimation with 4-D Automotive Radar](https://mlsanity.com/paper/2203.01137/) (2022)
- [milliFlow: Scene Flow Estimation on mmWave Radar Point Cloud for Human Motion Sensing](https://mlsanity.com/paper/2306.17010/) (2023)
- [Real-time People Tracking and Identification from Sparse mm-Wave Radar Point-clouds](https://mlsanity.com/paper/2105.11368/) (2021)
- [Saving Human Lives: What Complexity Science and Information Systems can Contribute](https://mlsanity.com/paper/1402.7011/) (2014)
- [UNIST: Unpaired Neural Implicit Shape Translation Network](https://mlsanity.com/paper/2112.05381/) (2021)
- [DeepCount: Crowd Counting with WiFi via Deep Learning](https://mlsanity.com/paper/1903.05316/) (2019)
- [Estimating indoor occupancy through low-cost BLE devices](https://mlsanity.com/paper/2102.03351/) (2021)
- [Advanced Millimeter-Wave Radar System for Real-Time Multiple-Human Tracking and Fall Detection](https://mlsanity.com/paper/2403.05634/) (2024)
- [ORACLE: Occlusion-Resilient and Self-Calibrating mmWave Radar Network for People Tracking](https://mlsanity.com/paper/2208.14199/) (2022)
- [Crowd Size Estimation for Non-Uniform Spatial Distributions with mmWave Radar](https://mlsanity.com/paper/2410.09200/) (2024)
- [Joint Multitarget Detection and Tracking with mmWave Radar](https://mlsanity.com/paper/2412.17211/) (2024)

## Abstract (arXiv)

In this paper, we present mmFlux: a novel framework for extracting underlying crowd motion patterns and inferring crowd semantics using mmWave radar. First, our proposed signal processing pipeline combines optical flow estimation concepts from vision with novel statistical and morphological noise filtering. This approach generates high-fidelity mmWave flow fields-compact 2D vector representations of crowd motion. We then introduce a novel approach that transforms these fields into directed geometric graphs. In these graphs, edges capture dominant flow currents, vertices mark crowd splitting or merging, and flow distribution is quantified across edges. Finally, we show that analyzing the local Jacobian and computing the corresponding curl and divergence enables extraction of key crowd semantics for both structured and diffused crowds. We conduct 21 experiments on crowds of up to 20 people across 3 areas, using commodity mmWave radar. Our framework achieves high-fidelity graph reconstruction of the underlying flow structure, even for complex crowd patterns, demonstrating strong spatial alignment and precise quantitative characterization of flow split ratios. Finally, our curl and divergence analysis accurately infers key crowd semantics, e.g., abrupt turns, boundaries where flow directions shift, dispersions, and gatherings. Overall, these findings validate mmFlux, underscoring its potential for various crowd analytics applications.